Font Size: a A A

System Design And Implementation For Precursor Chemical Trading Behavior Analysis

Posted on:2019-06-29Degree:MasterType:Thesis
Country:ChinaCandidate:R J ZhangFull Text:PDF
GTID:2348330542955206Subject:Computer technology
Abstract/Summary:PDF Full Text Request
The Information System On Supervision And Management Of Precursor Chemicals produced by Hebei SBS Innovation Technology Co.,Ltd.helped the regulatory department strengthen the supervision of precursor chemicals and related enterprises.But existing system has not meet the regulatory requirements of studying the relationship among enterprises,finding the suspicious enterprises,analyzing the flow of precursor chemicals,making decision and prediction of trend.Therefore,the regulatory department was pressed for an analysis system,which could analyze the relationship of the purchase among enterprises and the flow of precursor chemicals timely.This thesis is based on the data warehouse of eight provinces and one city.To supervise the enterprise relationship and the flow of precursor chemicals,we extracted the trading behavior data of precursor chemicals,analyzed the statistical characteristics of enterprise relationship network,implemented the algorithm of analyzing the flow of precursor chemicals.Eventually,the Trading Behavior Analysis System for Precursor Chemicals was completed successfully.At present,the system has operated normally.The main work and innovation of this thesis are as follows:1.Construction of enterprise relationship network.Based on complex network theory,we extracted and sorted out the data of purchasing and transporting from the data warehouse.From the three dimensions of time,area and chemicals,we constructed the enterprise relationship network whose node represented a single enterprise and edge represented the enterprise relationship.2.Analysis of the statistical characteristics of enterprise relationship network.We calculated the value of the enterprise relationship network statistical characteristics by applying the NetworkX module of Python.And we realized the visualization by using ECharts.3.Division of the community structures of enterprise relationship network.This thesis successfully applied a community structure partition algorithm based on the improved maximum flow problem.And we could find the enterprise community in a close relationship to analyze the characteristics of enterprise trading behavior.4.Design and implementation of the global search algorithm and the local search algorithm.This thesis first proposed and implement the global search algorithm and the local search algorithm to analyze the flow of precursor chemicals.We extracted and organized all trading behavior data of chemicals from the data warehouse.By tracing the source upwards and downwards,we finally implemented the entire flow analysis process.
Keywords/Search Tags:complex networks, the enterprise relationship network, statistical features, community structure, flow analysis
PDF Full Text Request
Related items